#005 - Statistical Analysis for Multiple Stocks¶

On this code, we perform some statistical analisys, from data obtained from #004. Finaly we plot some graphics as correlation, scatterplots and histograms.

For future: automate lists to run any given stocks

Link to Seaborn Documentation: https://seaborn.pydata.org/

Install Libs. (remove comments '#' if need to install the libraries)

In [1]:
#    !pip install pandas
#    !pip install pandas-datareader
#    !pip install numpy
#    !pip install yfinance
#    !pip install datetime
#    !pip install plotly_express
#    !pip install seaborn
#    !pip install matplotlib
In [2]:
#import Libraries
import pandas as pd
from pandas_datareader import data as pdr
import numpy as np
import yfinance as yf
import datetime as dt
import plotly.express as px
import plotly.graph_objects as go
import matplotlib.pyplot as plt
import seaborn as sns

Define functions¶

In [3]:
# Function to scale stock prices based on their initial starting price
# The objective of this function is to set all prices to start at a value of 1 
def price_scaling(raw_prices_df):
    scaled_prices_df = raw_prices_df.copy()
    for i in raw_prices_df.columns[0:]:
          scaled_prices_df[i] = raw_prices_df[i]/raw_prices_df[i][0]
    return scaled_prices_df
In [4]:
# Define a function using Plotly Express
def plotly_data(df, title):  
    
    # Create figure
    fig = go.Figure()
   
    # Set title
    fig.update_layout(title_text = title) 
    
    # For loop that plots all stock prices in the pandas dataframe df
    
    for i in df.columns[0:]:
        # Add range slider
        #fig.update_layout(xaxis=dict(rangeselector = dict(buttons=list([dict(count=1, label="1m", step="month", stepmode="backward"), dict(count=6, label="6m", step="month", stepmode="backward"), dict(count=1, label="YTD", step="year", stepmode="todate"), dict(count=1, label="1y", step="year", stepmode="backward"), dict(step="all")])), rangeslider=dict( visible=True), type="date"))
        # Add line graph
        fig.add_scatter(x = df.index, y = df[i], name = i)
        # Update Layout
        fig.update_layout({'plot_bgcolor': "white"})
        #fig.update_traces(line_width = 3)
    
    fig.show()
    
# Define a function using Plotly Express, changes axis y to logarithm scale
def log_plotly_data(df, title):  
    
    # Create figure
    fig = go.Figure()
   
    # Set title
    fig.update_layout(title_text = title) 
    
    # For loop that plots all stock prices in the pandas dataframe df
    
    for i in df.columns[0:]:
        # Add range slider
        #fig.update_layout(xaxis=dict(rangeselector = dict(buttons=list([dict(count=1, label="1m", step="month", stepmode="backward"), dict(count=6, label="6m", step="month", stepmode="backward"), dict(count=1, label="YTD", step="year", stepmode="todate"), dict(count=1, label="1y", step="year", stepmode="backward"), dict(step="all")])), rangeslider=dict( visible=True), type="date"))
        # Add line graph
        fig.add_scatter(x = df.index, y = df[i], name = i)
        # Update Layout
        fig.update_layout({'plot_bgcolor': "white"})
    
    #changes y to logarithm scale
    fig.update_yaxes(type="log")
    fig.show()

5.1 Import and Organize Data¶

In [5]:
# input stocks and organize into a List Object, 
tickers = [item for item in input("Enter the stock tickers, for portfolio (space them only) : ").split()]   
#   PETR4.SA VALE3.SA TAEE11.SA ITSA4.SA WEGE3.SA
#   AAPL AMZN GOOG META MSFT NVDA TSLA

yf.pdr_override()

# define a data range, 
end = dt.datetime.now()
start = end - dt.timedelta(days = 365*10)

Enter the stock tickers, for portfolio (space them only) : AAPL AMZN GOOG META MSFT NVDA TSLA
In [6]:
#obtain data from YFinance
df = pdr.get_data_yahoo(tickers, start = start, end = end)
df

[*********************100%%**********************]  7 of 7 completed
Out[6]:
Adj Close Close ... Open Volume
AAPL AMZN GOOG META MSFT NVDA TSLA AAPL AMZN GOOG ... MSFT NVDA TSLA AAPL AMZN GOOG META MSFT NVDA TSLA
Date
2013-12-13 17.359430 19.212000 26.420719 53.320000 30.895048 3.542601 9.843333 19.801071 19.212000 26.420719 ... 37.419998 3.785000 9.870000 332822000 60500000 86820208 82641000 40066100 18458800 158878500
2013-12-16 17.455557 19.448500 26.724331 53.810001 31.063459 3.547319 9.862667 19.910713 19.448500 26.724331 ... 36.730000 3.757500 9.898667 282592800 45034000 64320188 85119000 31734200 23762000 100129500
2013-12-17 17.376966 19.382500 26.646622 54.860001 30.751888 3.563828 10.164000 19.821072 19.382500 26.646622 ... 36.939999 3.757500 9.838667 229902400 46878000 61658247 78751000 45687700 27410400 157425000
2013-12-18 17.244839 19.798000 27.017483 55.570000 30.802433 3.613359 9.865333 19.670357 19.798000 27.017483 ... 36.360001 3.785000 10.149333 565863200 69782000 88743390 76003000 63192100 42623200 173728500
2013-12-19 17.047266 19.759501 27.054094 55.049999 30.524549 3.627511 9.381333 19.445000 19.759501 27.054094 ... 36.509998 3.825000 9.793333 320308800 48544000 66877738 89753200 34160100 25854400 191100000
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
2023-12-04 189.429993 144.839996 130.630005 320.019989 369.140015 455.059998 235.580002 189.429993 144.839996 130.630005 ... 369.100006 460.769989 235.750000 43389500 48294200 24117100 19037100 32063300 43754300 104099800
2023-12-05 193.419998 146.880005 132.389999 318.290009 372.519989 465.660004 238.720001 193.419998 146.880005 132.389999 ... 366.450012 454.660004 233.869995 66628400 46822400 19235100 16952100 23065000 37171800 137971100
2023-12-06 192.320007 144.520004 131.429993 317.450012 368.799988 455.029999 239.369995 192.320007 144.520004 131.429993 ... 373.540009 472.149994 242.919998 41089700 39679000 16360600 11294300 21182100 38059000 126436200
2023-12-07 194.270004 146.880005 138.449997 326.589996 370.950012 465.959991 242.639999 194.270004 146.880005 138.449997 ... 368.230011 457.000000 241.550003 47477700 52352800 38419400 15905100 23118900 35082300 107142300
2023-12-08 195.710007 147.419998 136.639999 332.750000 374.230011 475.059998 243.839996 195.710007 147.419998 136.639999 ... 369.200012 465.950012 240.270004 53377300 41858000 22990900 14077500 20144800 35880300 102980100

2514 rows × 42 columns

In [7]:
#select Ajusted Close Prices
Adj_Close_df = df['Adj Close']

#select Volume
volume_df = df['Volume']

#calculate percentage daily return
p_change_df = Adj_Close_df.pct_change() * 100
p_change_df.replace(np.nan, 0, inplace = True)

#Scale close prices
scaling_df = price_scaling(Adj_Close_df)
scaling_df
Out[7]:
AAPL AMZN GOOG META MSFT NVDA TSLA
Date
2013-12-13 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000
2013-12-16 1.005537 1.012310 1.011491 1.009190 1.005451 1.001332 1.001964
2013-12-17 1.001010 1.008875 1.008550 1.028882 0.995366 1.005992 1.032577
2013-12-18 0.993399 1.030502 1.022587 1.042198 0.997002 1.019973 1.002235
2013-12-19 0.982018 1.028498 1.023973 1.032446 0.988008 1.023968 0.953065
... ... ... ... ... ... ... ...
2023-12-04 10.912224 7.539038 4.944226 6.001875 11.948194 128.453629 23.932950
2023-12-05 11.142071 7.645222 5.010840 5.969430 12.057595 131.445782 24.251948
2023-12-06 11.078705 7.522382 4.974505 5.953676 11.937188 128.445161 24.317981
2023-12-07 11.191036 7.645222 5.240205 6.125094 12.006779 131.530462 24.650186
2023-12-08 11.273988 7.673329 5.171699 6.240623 12.112945 134.099197 24.772096

2514 rows × 7 columns

Verify null values ocurrence and analize data¶

In [8]:
Adj_Close_df.describe().round(2)
Out[8]:
AAPL AMZN GOOG META MSFT NVDA TSLA
count 2514.00 2514.00 2514.00 2514.00 2514.00 2514.00 2514.00
mean 75.04 83.73 68.46 176.61 142.93 97.77 92.80
std 56.19 51.37 37.06 80.66 101.45 114.57 108.06
min 15.65 14.35 24.56 53.32 29.46 3.54 9.29
25% 27.11 35.88 37.40 114.51 49.13 11.47 15.69
50% 45.67 86.16 56.62 169.56 103.12 51.24 21.71
75% 131.62 125.98 99.17 217.45 238.21 142.91 196.76
max 195.93 186.57 150.71 382.18 382.70 504.05 409.97
In [9]:
p_change_df.describe().round(2)
Out[9]:
AAPL AMZN GOOG META MSFT NVDA TSLA
count 2514.00 2514.00 2514.00 2514.00 2514.00 2514.00 2514.00
mean 0.11 0.10 0.08 0.10 0.11 0.24 0.19
std 1.79 2.09 1.76 2.36 1.71 2.92 3.51
min -12.86 -14.05 -11.10 -26.39 -14.74 -18.76 -21.06
25% -0.72 -0.86 -0.71 -0.95 -0.68 -1.14 -1.53
50% 0.09 0.11 0.08 0.11 0.08 0.23 0.12
75% 1.02 1.13 0.93 1.23 0.98 1.63 1.87
max 11.98 14.13 16.05 23.28 14.22 29.81 19.89
In [10]:
scaling_df.describe().round(2)
Out[10]:
AAPL AMZN GOOG META MSFT NVDA TSLA
count 2514.00 2514.00 2514.00 2514.00 2514.00 2514.00 2514.00
mean 4.32 4.36 2.59 3.31 4.63 27.60 9.43
std 3.24 2.67 1.40 1.51 3.28 32.34 10.98
min 0.90 0.75 0.93 1.00 0.95 1.00 0.94
25% 1.56 1.87 1.42 2.15 1.59 3.24 1.59
50% 2.63 4.48 2.14 3.18 3.34 14.46 2.21
75% 7.58 6.56 3.75 4.08 7.71 40.34 19.99
max 11.29 9.71 5.70 7.17 12.39 142.28 41.65

5.2 Plotting Pair Data¶

In [11]:
#generate variables for stocks list
#need to copy the output and paste on next cell

i = range(len(tickers))
for n in i:
    j = p_change_df.iloc[:, n] 
    tk = tickers[n]
    print(tickers[n], '= p_change_df["{}"]'.format(tk))

AAPL = p_change_df["AAPL"]
AMZN = p_change_df["AMZN"]
GOOG = p_change_df["GOOG"]
META = p_change_df["META"]
MSFT = p_change_df["MSFT"]
NVDA = p_change_df["NVDA"]
TSLA = p_change_df["TSLA"]
In [12]:
# have to change this for every run

AAPL = p_change_df["AAPL"]
AMZN = p_change_df["AMZN"]
GOOG = p_change_df["GOOG"]
META = p_change_df["META"]
MSFT = p_change_df["MSFT"]
NVDA = p_change_df["NVDA"]
TSLA = p_change_df["TSLA"]
In [13]:
#Set Parameters

x = GOOG
y = META
ulim = 5
dlim = - ulim
In [14]:
# a joinplot with percentage daily returns, date is constant 1 day
sns.jointplot(data=x, color = 'r', )
sns.jointplot(data=y, color = 'r', )
Out[14]:
<seaborn.axisgrid.JointGrid at 0x2b068c5d2d0>
In [15]:
#joint 2 stocks into a scatter

sns.jointplot(x=x, y=y, xlim = {dlim, ulim}, ylim = {dlim, ulim})
Out[15]:
<seaborn.axisgrid.JointGrid at 0x2b069373590>
In [16]:
#Another plot types

sns.jointplot(x=x, y=y, kind="hist", xlim = {dlim, ulim}, ylim = {dlim, ulim})
sns.jointplot(x=x, y=y, kind="hex", xlim = {dlim, ulim}, ylim = {dlim, ulim})
sns.jointplot(x=x, y=y, kind="kde", xlim = {dlim, ulim}, ylim = {dlim, ulim})
sns.jointplot(x=x, y=y, kind="reg", line_kws=dict(color="r"), marker = "^", xlim = {dlim, ulim}, ylim = {dlim, ulim})
Out[16]:
<seaborn.axisgrid.JointGrid at 0x2b06a4664d0>
In [17]:
# Add more layers into the plot

g = sns.jointplot(x=x, y=y, marker = "v", xlim = {dlim, ulim}, ylim = {dlim, ulim})
g.plot_joint(sns.kdeplot, color="r", zorder=0, levels=8)
g.plot_marginals(sns.rugplot, color=".5", height=-.15, clip_on=False)
Out[17]:
<seaborn.axisgrid.JointGrid at 0x2b06a523a90>
In [18]:
#perform a linear regression
    
sns.regplot(x=x, y=y
            # polinomial order to ajust regression
            , order = 1
            # confidence interval
            , ci=99
            #change color range saturation 0 black to 1 white
            , color=".6"
            # change color from regression line 'r' = red
            , line_kws=dict(color="r")
            # Fit a robust regression to downweight the influence of outliers
            #, robust=True
            
            )
Out[18]:
<Axes: xlabel='GOOG', ylabel='META'>

5.3 Plot Entire Data Stocks¶

In [19]:
log_plotly_data(scaling_df, "Scaled Prices")
In [20]:
# Plot histograms for stocks daily returns using plotly express
fig = px.histogram(p_change_df)
fig.update_layout({'plot_bgcolor': "white"})
In [21]:
# Plot a heatmap showing the correlations between daily returns

plt.figure(figsize = (10, 8))
sns.heatmap(p_change_df.corr(), annot = True, vmin= -1, vmax= 1, cmap= "Spectral")
Out[21]:
<Axes: >
In [22]:
# Plot the Pairplot between stocks daily returns
sns.pairplot(p_change_df)

C:\Users\marqu\anaconda3\Lib\site-packages\seaborn\axisgrid.py:118: UserWarning:

The figure layout has changed to tight
Out[22]:
<seaborn.axisgrid.PairGrid at 0x2b06c3abbd0>
In [23]:
# Plot the Pairplot between stocks daily returns

g = sns.PairGrid(p_change_df, diag_sharey=False)
g.map_upper(sns.kdeplot)
g.map_lower(sns.regplot,line_kws=dict(color="r"), marker = "^")
g.map_diag(sns.histplot, kde=True, lw=2, )
Out[23]:
<seaborn.axisgrid.PairGrid at 0x2b070d17dd0>